Epididymal Dysfunction Initiated by the Expression of Simian Virus 40 ...

0888-8809/02/$15.00/0 Printed in U.S.A.

Molecular Endocrinology 16(11):2603–2617 Copyright © 2002 by The Endocrine Society doi: 10.1210/me.2002-0100

Epididymal Dysfunction Initiated by the Expression of Simian Virus 40 T-Antigen Leads to Angulated Sperm Flagella and Infertility in Transgenic Mice PETRA SIPILA¨, TREVOR G. COOPER, CHING-HEI YEUNG, MIKA MUSTONEN, JENNI PENTTINEN, JOE¨L DREVET, ILPO HUHTANIEMI, AND MATTI POUTANEN Department of Physiology, Institute of Biomedicine (P.S., M.M., J.P., I.H., M.P.), and Turku Graduate School of Biomedical Sciences (P.S., J.P.), University of Turku, FIN-20520 Turku, Finland; Institute of Reproductive Medicine of the University (T.G.C., C.-H.Y.), D-48129 Mu¨nster, Germany; and Reproduction and Development Research Group (J.D.), Blaise Pascal University, Centre National de la Recherche Scientifique Unite´ Mixte de Recherche, F-63177 Aubie`re, France We have generated two transgenic mouse lines (GPX5-Tag1 and GPX5-Tag2) by expressing the Simian virus 40 large and small T-antigens under a 5-kb promoter of the murine glutathione peroxidase 5 (GPX5) gene. In GPX5-Tag1 mice, with a high level of T-antigen expression, severe dysplasia was found in the epididymis and seminal vesicles. These mice also developed adrenal and prostate tumors, and spermatogenesis was disrupted. In GPX5-Tag2 mice, with a lower level of T-antigen expression, the only histological change was the slightly hyperplastic epithelium in the initial segment of the epididymidis and in the seminal vesicles. Despite normal mating behavior, these mice were infertile. The most conspicuous feature of the

sperm was angulation of the flagellum, which appeared during epididymal transit, probably due to the observed reduction in the osmotic pressure of cauda epididymidal fluid. The angulation did not affect the motility or kinematic parameters of the sperm, but the sperm were also incapable of fertilization in vitro. The lack of expression of several genes specific for the initial segment suggests that in the GPX5-Tag2 mice the transgene expression brings about a differentiation arrest in this part of epididymis. This novel mouse line provides a model for epididymal dysfunction leading to defects in posttesticular sperm maturation and infertility. (Molecular Endocrinology 16: 2603–2617, 2002)

T

mal epithelium by targeting expression of the Simian Virus 40 large T-antigen (SV40 Tag) into the epididymis of transgenic mice by using the murine GPX5 promoter. Recently, the 5⬘-flanking sequences of two epididymis-specific proteins have been shown to direct the transgene expression into the epididymis, namely those of murine epididymal retinoic acid-binding protein (mE-RABP; Ref. 5) and murine glutathione peroxidase 5 (GPX5; Ref. 6). GPX5 is a well characterized epididymal secretory protein that belongs to the glutathione peroxidase family (7, 8). The expression of GPX5 in the caput epididymidis is under the control of androgens (7, 9), and we have shown that the 5.0-kb 5⬘ flanking region of the mouse GPX5 gene is suitable for directing transgene expression to the caput epididymidis (6). The function of GPX5 is not totally understood, but it has been shown to bind to epididymal spermatozoa (10, 11) at the acrosomal region, and therefore it has been suggested that GPX5 could protect sperm membranes from oxidative damage (8, 12, 13). SV40 Tag is able to transform cells (14), and it is widely used in studies on genetically targeted tumorigenesis owing to its capacity of transforming even well differentiated cell types (15–17). The exact mechanisms for immortalization and transformation by SV40 Tag are not yet known, but several immortaliza-

HE MATURATION OF mammalian spermatozoa is acquired in the epididymis, where spermatozoa gain their motility and fertilizing capacity. A number of morphological and biochemical changes occur in spermatozoa during their transit through the epididymis, and the most consistent morphological change is the migration of the cytoplasmic droplet from the neck region of the flagellum to the end of the midpiece (for review, see Ref. 1). The epididymis plays an active role in sperm development, and sperm maturation is dependent on the unique luminal environment of the epididymis, including specific proteins synthesized and secreted by the epididymal epithelium (2, 3). Although several epididymis-specific secretory proteins have been identified, little is known about the sperm maturation events in the epididymis (4). Our aim has been to understand epididymal function further to develop novel strategies for nonhormonal male contraception. In this study, we have taken the approach of immortalizing and/or disrupting the function of the epididyAbbreviations: CRES, Cystatin-related epididymal spermatogenic; DTT, dithiothreitol; EGFP, enhanced green fluorescent protein; GPX5, glutathione peroxidase 5; HE6, human epididymal 6; mEP17, mouse epididymal protein 17; SV40 Tag, Simian virus 40 large T-antigen; TUNEL, terminal uridine deoxynucleotidyl nick end labeling; WT, wild-type.

2603

2604 Mol Endocrinol, November 2002, 16(11):2603–2617

Sipila¨ et al. • Epididymal Dysfunction with Infertility in Mice

tion pathways are likely to be involved (for review, see Ref. 18). It has been suggested that binding of SV40 Tag to two cell cycle regulatory proteins, p53 and retinoblastoma susceptibility protein, is needed for transformation and immortalization (19), but p53-independent pathways have also been reported (18, 19). Furthermore, it has been shown that in addition to tumorigenesis, T-antigen expression in transgenic mice can cause apoptosis, for example in the mammary gland (20). Epididymal epithelial cells are highly differentiated and possess different characteristic features in different regions of epididymis. As shown in the present study, a high level of expression of SV40 Tag in the epididymis leads to severe dysplasia, whereas a low level of expression was found to cause epididymal dysfunction, resulting in a maturation defect of the sperm leading to angulated sperm flagella and infertility.

further detected in the prostate tumors formed. In contrast, Northern blot analysis of the GPX5-Tag2 males indicated that the transgene was expressed exclusively in the epididymis at both age groups analyzed (Fig. 1C). In addition to that found by the Northern analysis, RT-PCR analysis also showed a weak expression in the testis in both of the mouse lines and in vas deferens and seminal vesicles in the GPX5-Tag2 mice.

RESULTS Generation of GPX5-Tag Transgenic Mice After pronucleus injection and embryo transfer, six GPX5-Tag founder mice were obtained (two males and four females). Transgene copy numbers were determined for each founder mouse by Southern blot analysis, and the amount of integrated copies of the GPX5-Tag transgene varied between 1 and 5 (data not shown). Two female founders with one copy of the transgene (020 and 049) transmitted the gene into the next generation in a Mendelian fashion, and two transgenic mouse lines (GPX5-Tag1 and GPX5-Tag2) were generated by breeding with the wild-type (WT) mice. All of the male mice in both GPX5-Tag1 and GPX5Tag2 lines were proven to be infertile. Hence, transgenic lines could be established only by mating the GPX5-Tag1 and GPX5-Tag2 females with the WT males. GPX5-Tag1 females typically died at the age of 3–6 months, and the histo-pathological changes found in the females were highly variable. The most constant finding was that they developed adrenal tumors. In addition, tumors were infrequently detected in the pituitary gland and uterus. By contrast, the GPX5-Tag2 female mice had a normal life span, and no tumor formation was found in them. Expression of the GPX5-Tag mRNA in GPX5-Tag1 and GPX5-Tag2 Male Mice Northern blot and RT-PCR analyses were used to identify the transgene expression in GPX5-Tag1 and GPX5-Tag2 males at the ages of 50 d and 4 months. In the GPX5-Tag1 males, transgene expression was found in the epididymis, seminal vesicles, vas deferens, and adrenal gland at the age of 50 d (Fig. 1B), and at the age of 4 months the transgene mRNA was

Macroscopic and Histological Characteristics of the GPX5-Tag1 and GPX5-Tag2 Male Mice In line with the strong SV40 Tag mRNA expression, macroscopic evaluation of the GPX5-Tag1 males at the age of 4 months revealed that their seminal vesicles were highly enlarged, with a 20-fold increase in weight. In addition, the weights of the adrenals and prostate block were enlarged by 96-fold and 16-fold, respectively, in the 4-month-old GPX5-Tag1 males (Table 1 and Fig. 2). However, the abnormal growth of these tissues took place only after puberty because no weight gain in any of the tissues was observed at the age of 50 d (Table 1). The testis size was also normal in the 50-d-old GPX5-Tag1 mice. However, after puberty the testis size did not grow normally, and by the age of 4 months the testis weight was 58% lower than in age-matched controls. Macroscopic changes in the genital tract of GPX5-Tag2 males were much milder. The only notable macroscopic difference was that the strong vascularization typical for the initial segment of caput epididymidis was not evident. No differences were found in weights of the seminal vesicles, prostate block, or adrenal gland compared with WT males (Table 1 and Fig. 2). At 4 months of age, however, the weights of epididymides and testes were both 43% lower than in WT mice, and the sizes were similar to those measured in 50- to 60-d-old mice (Table 1). Furthermore, no signs of tumor formation were seen in any of the GPX5-Tag2 mice when followed up to 10 months of age. Histological evaluation of the affected organs of the GPX5-Tag1 and GPX5-Tag2 male mice was performed at the ages of 2 and 4 months. At the age of 2 months, seminal vesicles of the GPX5-Tag1 were already dysplastic (Fig. 3B), and at the age of 4 months severe dysplasia was found in the highly enlarged seminal vesicles. At the age of 2 months, prostate dysplasia was detected; this developed to a highgrade prostate cancer in all prostate lobes by the age of 4 months, which was also indicated by the growth of prostate size. Furthermore, the enlarged adrenals showed tumors originating from the zona fasciculata of the adrenal cortex (Fig. 3H). In the epididymis, severe dysplasia with no signs of invasion of the tumor cells was already found in the caput, corpus, and cauda at the age of 2 months, and epididymal duct was occluded at corpus and cauda regions (Fig. 4, B, E, H, and K). At the same age, the spermatogenesis of the GPX5-Tag1 males was disrupted, and normal


Mol Endocrinol, November 2002, 16(11):2603–2617

2605

Fig. 1. Transgene Expression in GPX5-Tag1 and GPX5-Tag2 Mouse Lines A, Schematic map of the GPX5-Tag transgene construct used in the transgenic mice production. A 5-kb-long murine GPX5 promoter fragment was ligated in front of a 2.7-kb fragment of the SV40 early genes in the pGEM-7Zf vector, and the GPX5-Tag transgene created was cleaved from the vector with XhoI and SacI digestions. B, Northern blot analysis of the GPX5-Tag transgene expression in GPX5-Tag1. C, GPX5-Tag2 male mouse at the age of 50 d. In the GPX5-Tag1 males, transgene expression was found in the epididymis, seminal vesicles, vas deferens, and adrenal gland. In contrast, in the GPX5-Tag2 males, transgene expression was detected only in the epididymis.

spermatogenesis was present only in a few tubules (Fig. 3E). These changes found in the epididymis and testis were even more intense at the age of 4 months. Except for the testis, in all affected tissues SV40 Tag expression was confirmed by immunohistochemistry (Fig. 5, B, E, H, J, K, and L). In the GPX5-Tag2 males with a lower level of transgene mRNA in the epididymis, no severe histological changes at light microscopic level were observed. However, at the age of 2 months, the initial segment of the transgenic epididymis displayed an apparently hyperplastic epithelium (Fig. 4, C and F) compared with the WT organs (Fig. 4, A and D), resulting in stellate luminal profiles. The epithelium of the seminal vesicles was also slightly hyperplastic (Fig. 3C) compared with the WT males (Fig. 3A). No obvious histological

changes were present in the testis (Fig. 3F), the adrenal gland (Fig. 3I), or the other tissues analyzed by the age of 4 months. Immunohistochemical analysis revealed that similar to that found in GPX5-Tag1 mice, Tag was expressed in all epididymal regions in the SV40 GPX5-Tag2 males (Fig. 5, C, F, and I), whereas no staining was found in testis (data not shown). Terminal Uridine Deoxynucleotidyl Nick End Labeling (TUNEL) Staining To study whether increased apoptosis rate was the cause of decreased testicular and epididymal size in GPX5-Tag2 males, TUNEL staining was performed. In the GPX5-Tag2 initial segment, the number of apoptotic cells per independent microscopic frame was sig-



Table 1. Organ Weightsa of the Reproductive Tissues in the WT and GPX5-Tag1 and -Tag2 Mice Animal Line

50 d of Age WT GPX5-Tag1 GPX5-Tag2 4 Months of Age WT GPX5-Tag1 GPX5-Tag2

Epididymis

Testis

Seminal Vesicles

46 (⫾23) 46 (⫾21) 36 (⫾10)

99 (⫾39) 99 (⫾23) 93 (⫾8)

195 (⫾39) 148 (⫾23)b 202 (⫾82)

74 (⫾27) 82 (⫾84) 42 (⫾42)g

140 (⫾20) 59 (⫾20)c 80 (⫾22)h

341 (⫾56) 7088 (⫾3605)d 271 (⫾42)

Adrenal Glands

3.73 (⫾21) 3.28 (⫾0.1) 3.41 (⫾0.3) 2.0 (⫾0.6) 360 (⫾352)e 3.23 (⫾0.5)

Prostate

47 (⫾8) 41 (⫾8) 34 (⫾4) 34 (⫾8) 741 (⫾861)f 53 (⫾27)

Mean weight in milligrams (⫾ SD); n ⫽ 5–10 animals. GPX5-Tag1 differed significantly from GPX5-Tag2 (P ⬍ 0.05). c GPX5-Tag1 differed significantly from GPX5-Tag2 (P ⬍ 0.01) and WT (P ⬍ 0.001). d GPX5-Tag1 differed significantly from GPX5-Tag2 and WT (P ⬍ 0.01). e GPX5-Tag1 differed significantly from GPX5-Tag2 and WT (P ⬍ 0.001). f GPX5-Tag1 differed significantly from WT (P ⬍ 0.001) and GPX5-Tag2 (P ⬍ 0.01). g GPX5-Tag2 differed significantly from WT (P ⬍ 0.01) and GPX5-Tag1 (P ⬍ 0.001). h GPX5-Tag2 differed significantly from WT (P ⬍ 0.001). a

b

Fig. 2. Reproductive Tract and Adrenal Phenotype of the GPX5-Tag1, GPX5-Tag2, and WT Male Mice at the Age of 3 Months The macroscopic evaluation of GPX5-Tag1 male mice revealed highly enlarged seminal vesicles, prostate, and adrenal glands. In addition, testes were smaller than the testes of control animals. The genital tract of GPX5-Tag2 male mouse was macroscopically indistinguishable from that of the WT mouse. A, Adrenal gland; K, kidney; SV, seminal vesicle; VD, vas deferens; E, epididymis; P, prostate.

nificantly higher (13.4 ⫾ 1.93; P ⬍ 0.001; n ⫽ 7; Fig. 6A) as compared with the WT initial segment (1.1 ⫾ 0.51; Fig. 6B). Also, in the caput region the number of apoptotic cells was higher in GPX5-Tag2 males (3.0 ⫾ 1.0; P ⬍ 0.05) than in the WT males (0.7 ⫾ 0.29). By contrast, no difference was found in the number of apoptotic cells between the genotypes in corpus (WT, 0.7 ⫾ 0.42; GPX5-Tag2, 0.7 ⫾ 0.56) and cauda regions (WT, 0; GPX5-Tag2, 0.4 ⫾ 0.42). Furthermore, the number of apoptotic cells in the different spermatogenic stages (I–VI, VII–VIII, and IX–XII) was equal be-

tween WT (0.4 ⫾ 0.20, 0.6 ⫾ 0.41, and 1.1 ⫾ 0.24, respectively) and GPX5-Tag2 males (0.1 ⫾ 0.06, 0.4 ⫾ 0.11, and 0.8 ⫾ 0.27). Hormone Measurements At both 50 d and 4 months of age, no difference was found in serum testosterone concentrations between the GPX5-Tag1, GPX5-Tag2, and WT mice. Accordingly, the LH levels were similar between the different groups. In FSH, no difference was found at the age of



2607

Fig. 3. Histology of the Seminal Vesicles, Testis, and Adrenal Gland of the WT, GPX5-Tag1, and GPX5-Tag2 Mice at the Age of 2 Months A, D, and G, WT mice. There were no severe histological changes in the seminal vesicles (C), testis (F), or adrenal gland (I) of the GPX5-Tag2. In contrast, the epithelium of the seminal vesicles (B) was dysplastic in the GPX5-Tag1 males compared with WT mice, and spermatogenesis in the seminiferous tubules was disrupted (E). Furthermore, adrenal gland tumors were originating from the zona fasciculata of the adrenal cortex (H). AD, Adrenal gland; S, seminiferous tubules; SV, seminal vesicles; T, testis.

50 d, whereas at 4 months of age the FSH values were clearly reduced in GPX5-Tag1 males (19.9 ⫾ 3.1 ␮g/ liter; P ⬍ 0.001) and to a lesser extent in the GPX5Tag2 males (34.8 ⫾ 1.4 ␮g/liter; P ⬍ 0.01) compared with the WT males (39.7 ⫾ 3.6 ␮g/liter). At the age of 4 months, serum inhibin B levels were slightly higher in GPX5-Tag1 and GPX5-Tag2 than in WT male mice, but the difference did not reach statistical significance (results not shown). Reproductive Performance of the GPX5-Tag2 Mice Because of the severely altered testis structure, the lack of sperm, and the drastic structural changes in the epithelium in the epididymis and seminal vesicles, it was not surprising to find that all of the GPX5-Tag1 males were infertile. However, all of the GPX5-Tag2 males analyzed were also infertile, although histological analyses of the reproductive tissues of these male mice did not show any obvious reason for the infertil-

ity. The reproductive performance of the GPX5-Tag2 males was therefore examined further. Because the mice possessed normal mating behavior (copulatory plugs in 18 of 18 females studied), and the uterus was full of sperm after mating, anejaculation was obviously not the reason for the infertility observed in the GPX5Tag2 males. The ability of the sperm to fertilize oocytes in vivo was studied next. After mating the GPX5Tag2 males with WT females, neither sperm nor fertilized oocytes were found in the oviducts of the females (Table 2). Hence, it was likely that the spermatozoa produced by the GPX5-Tag2 males were unable to reach the oviduct in normal matings. As a control, the oocytes in the females mated with WT males were analyzed, and 96% of the oocytes were fertilized in vivo (Table 2). Further fertilization tests showed that the sperm obtained from the GPX5-Tag2 mice were also incapable of fertilization of the oocytes in vitro. When WT oocytes were incubated with sperm obtained from GPX5-Tag2



Fig. 4. Histology of the Initial Segment, Corpus, and Cauda Epididymidis of WT, GPX5-Tag1, and GPX5-Tag2 Male Mice at the Age of 2 Months A, D, G, and J, WT mice. Severe dysplasia was found in the initial segment (B, E), corpus (H), and cauda (K) epididymidis of GPX5-Tag1 male mice, with occlusions present in corpus and cauda epididymidis. Only detectable changes in the GPX5-Tag2 epididymis were hyperplastic epithelium and stellate luminal profiles in the initial segment (C, F). Corpus (I) and cauda (L) epididymides were histologically normal. Ca, Cauda; Co, corpus; IS, initial segment; St, stellate.

mice, no normal fertilization and embryo development was detected (Table 2). In contrast, WT sperm fertilized 22% of oocytes and 15% of embryos developed until the blastocyst stage in vitro (Table 2). As a control, oocytes from WT females from CBAxC57BL/6 strain were also used. Results were similar; none of the oocytes were fertilized when incubated with GPX5Tag2 sperm, and when WT sperm was used, 22% of embryos developed into the blastocyst stage.

Sperm Motility Spermatozoa from the cauda epididymidis of both GPX5-Tag2 and WT males were equally motile upon release in two test media used, and there was no difference in the maintenance of motility over the 2.5-h follow-up period. Because kinematic parameters were uninfluenced by the substrates, data from the two media were pooled. Although the percentage of motile



2609

Fig. 5. Immunohistochemical Staining of SV40 Tag in the Different Regions of the Epididymis Initial segment (IS), corpus (Co), and cauda (Ca) of WT (A, D, G), GPX5-Tag1 (B, E, H), and GPX5-Tag2 (C, F, I) mice, and adrenal gland (J), seminal vesicle (K), and prostate (L) of GPX5-Tag1 mice.

sperm from the cauda epididymidis was not different between GPX5-Tag2 and WT mice at either time point, there were marked differences in the flagella forms (Fig. 7). The majority of motile WT sperm had straight flagella, whereas most motile sperm from GPX5-Tag2 mice had hairpin bends (Fig. 7). The slight reductions in mean values of curvilinear velocity (micrometers per second), average path velocity (micrometers per second), straight-line velocity (micrometers per second), linearity (percentage), and amplitude of lateral head displacement (micrometers) of sperm from GPX5Tag2 mice did not reach statistical significance.

Sperm Morphology From the fixed epididymides, it was evident that the percentage of sperm with straight flagella decreased continuously as they migrated through the duct of GPX5-Tag2 males (Fig. 8). A particularly steep decrease was observed between epididymal regions 5 and 6 (Fig. 8C). As the straight forms declined, those with hairpin bends increased, whereas those with small or large bends remained constant (20%). By contrast, sperm from the WT mice were largely straight throughout the epididymis (Fig. 8C). The sperm from



mal region (percentage of droplet-bearing spermatozoa in the caput, corpus, and cauda of WT mice, 81.7 ⫾ 4.3, 94.6 ⫾ 1.5, and 85.1 ⫾ 3.4, respectively; and in GPX5-Tag2 mice, 81.6 ⫾ 4.2, 93.7 ⫾ 2.1, and 95.9 ⫾ 1.5, respectively), and Triton X-100 effectively removed them all. Incubation of intact GPX5-Tag2 caudal sperm in the reducing agent dithiothreitol (DTT) significantly reduced the percentage of sperm in hairpin bends and increased those with acute flagellar bends, but with no change in the total percentage of angulated forms. After demembranation of sperm incubated in DTT, the percentage of cauda sperm with hairpin bends was reduced, and the percentages of those with slight angulation and straight flagella was increased. Osmotic Pressure of Luminal Contents The osmotic pressure of the cauda epididymal contents from WT and GPX5-Tag2 males was also measured, and it was found to be significantly higher in WT (mean ⫾ SEM, 396 ⫾ 8 mmol/kg; n ⫽ 8) than in the GPX5-Tag2 mice (358 ⫾ 10 mmol/kg; n ⫽ 6). Epididymal Gene Expression in the GPX5-Tag2 Mouse Line

Fig. 6. TUNEL Staining of Apoptotic Cells in the Initial Segment of GPX5-Tag2 and WT Mice Microscopic frame from GPX5-Tag2 initial segment (A) and WT initial segment (B). Arrows indicate TUNEL-positive nuclei. The number of apoptotic cells was significantly higher (P ⬍ 0.001) in GPX5-Tag2 initial segment than in WT initial segment.

the WT and GPX5-Tag2 mice were then released into medium containing a detergent (Triton X-100) to analyze whether the bending was caused by membrane restraint. The majority of sperm (82–96%) from the WT caput, corpus, and cauda epididymidis and from the GPX5-Tag2 caput had straight flagella after immediate release into medium, whether or not they were demembranated for 1 min in 1% Triton X-100. By contrast, sperm from the corpus and cauda epididymidis of GPX5-Tag2 mice displayed lower percentages of straight forms (27% and 11%, respectively), whereas sperm with hairpin bends were prominent (50% in the corpus and 70% in the cauda). This data is thus consistent with the data found in fixed epididymides. Nevertheless, demembranation of corpus sperm resulted in fewer hairpin forms and more flagella with slight or acute angulation. For cauda sperm, hairpin forms were reduced in extent after demembranation, but they were still present (25%), and the percentage of those with acute angulation was increased. There were no differences between genotypes in the percentages of sperm that bore cytoplasmic droplets in any epididy-

To further characterize the epididymal dysfunction in GPX5-Tag2 mice, the expression of various epididymal and ion transporter genes (Table 3) was analyzed by RT-PCR and Northern blot analysis. There were no differences in the expression levels of 12 of 15 genes analyzed (data not shown). However, the mRNAs for two initial segment-specific genes, cystatin-related epididymal spermatogenic (CRES) and mouse epididymal protein 17 (mEP17), were found to be missing from GPX5-Tag2 mice (Fig. 9, A and B). Also, the mouse homolog for human epididymal 6 (HE6) was not expressed in the initial segment of GPX5-Tag2 mice, and the expression was lower in GPX5-Tag2 mice in distal caput and cauda compared with WT mice (Fig. 9C).

DISCUSSION The epididymis has been identified as the site where the essential posttesticular sperm maturation and storage occurs. Nevertheless, little is known about the process of sperm maturation and factors affecting it. Understanding the physiology of the epididymis could provide new possibilities for infertility treatment on one hand, and could suggest new strategies for the development of a new nonhormonal male contraceptive on the other. Our studies are directed toward understanding the physiology of the epididymis to develop novel strategies for male contraception based on inhibition of posttesticular sperm maturation in the epididymis. In the present study, we have generated epididymal



2611

Table 2. Fertilizing Capacity of the Sperm of WT and GPX5-Tag2 Male Mice in Vivo (A) and in Vitro (B) A In Vivo

No. of Males

Total No. of Copulatory Plugs

Total No. of Oocytes

Fertilized Oocytes in Vivo (%)

6 6

6 6

51 59

96 0

WT GPX5-Tag2 B

In Vitro

WT GPX5-Tag2 a

Fertilized Oocytes (%)

No. of Males

Total No. of Oocytes

2 Cell

4 Cell

8 Cell

Blastocyst

4 4

268 319

22 5a

16 0

16 0

15 0

Two-cell stages were found at d 2, but they did not develop further.

Fig. 7. Motility Parameters of Sperm Obtained from the Distal Cauda Epididymidis of WT and GPX5-Tag2 Males No differences in percentage of motile sperm (Mot) or kinematic parameters such as curvilinear velocity (VCL), average path velocity (VAP), straight-line velocity (VSL), linearity (LIN), or amplitude of head displacement (ALH) were observed. However, a significant difference in the form of the motile sperm was noted. About 80% of WT sperm showed a straight flagellum (Str), whereas only 10% of the sperm in GPX5-Tag2 males were straight. Conversely, about 80% of GPX5-Tag2 sperm swam with a hairpin bend (HPin), but only 20% of WT sperm were in this form. *, P ⬍ 0.05; the bars represent mean ⫾ SEM.

dysfunction in transgenic mice by introducing SV40 Tag under the control of GPX5 protein promoter. The GPX5 gene is expressed at high level in the caput epididymidis (21) but is also found at very low levels in various other tissues (6, 22). In accordance, expression of the 5-kb GPX5 promoter-driven enhanced green fluorescent protein (EGFP) transgene was recently found at high levels in the distal caput epididymidis, whereas few fluorescent cells were detected in the cauda region. A low level of EGFP expression was also detected in various other tissues (6). Hence, the SV40 Tag expression in all epididymal regions and the low-level expression outside the epididymis, found in the GPX5-Tag1 and -Tag2 mice, was not surprising, and as in the GPX5-EGFP transgenic mice, the cell

specificity of the expression was partly dependent on the integration site of the transgene. Several oncogenes have been used in studies aimed at developing animal models for tumor development. Some of these studies indicate that both the properties of the activated oncogene and the tissue context determine the final outcome, such as malignant transformation (23). There are several transgenic mouse models for tumor development in the prostate (15), testis (16), and adrenal gland (17). By contrast, in the epididymis only hyperplasia has been reported (23–25), and hyperplasia also seems to be a common feature in the seminal vesicles expressing an oncogene (24). These observations are in line with the mouse models generated in the present study. In the GPX5-Tag1 mice, tumors developed in the prostate and adrenal gland, whereas the seminal vesicles and epididymides were only hyperplastic with severe dysplasia. The disrupted spermatogenesis with involuted seminiferous tubules in GPX5-Tag1 mice is suggested to be caused by the occluded epididymal lumen at the corpus and caudal regions. Furthermore, it is also likely that the altered tubular structure in the initial segment and caput do not support the normal fluid flow in the epididymis. The conclusion of epididymal dysfunction leading to disrupted spermatogenesis is also supported by the fact that only a minor SV40 Tag expression was detected in the testis by RT-PCR and that in the earlier studies, no pathology has been observed in mice expressing SV40 Tag in haploid spermatids (26, 27). The phenotypic changes in GPX5-Tag2 males were less severe than in GPX5-Tag1 males. This is in line with a lower SV40 Tag expression in the GPX5-Tag2 males and the stricter expression of SV40 Tag in the epididymis. In the GPX5-Tag2 mouse line, one macroscopically notable difference was that the initial segment was lacking its normal reddish color, indicating a decrease in the vascularization. Increased vascularization is characteristic of the initial segment (28, 29) and is thought to underlie the high metabolic activity of this region (30). Decreased vascularity characterizes c-ros knockout mice, which lack the initial segment (31, 32). GPX5-Tag2 mice do have an initial segment, although light microscopy revealed stellate profiles in this region



Fig. 8. Morphology of Spermatozoa within the Epididymis of WT and GPX5-Tag2 Males Spermatozoa taken from the cauda epididymidis of WT mouse (A) and GPX5-Tag2 mouse (B). Arrows indicate hairpin flagella, and arrowheads indicate coiled flagella. C, WT males had predominantly (⬃90%) straight flagella (gray circles), and very few hairpin forms were observed (gray squares). A decline in straight forms characterized the GPX5-Tag2 males (black circles) and was mirrored by an increase in hairpin forms (black squares) reaching about 70% in the distal cauda epididymidis.

of the GPX5-Tag2 mice, which are not typical for the mouse. Such stellate profiles are characteristic of the initial segment of other species and are unlikely as such to contribute to an abnormal epididymal environment. The other difference between the GPX5-Tag2 and WT mice was the lower testis weight in the GPX5Tag2 mice at older age. An increased rate of apoptosis has been reported in many SV40 Tag mouse models, and some studies indicate that apoptosis has a role in regulating tumorigenesis (33–35), whereas other studies indicate that the expression of SV40 Tag directly causes apoptosis (20). In our GPX5-Tag2 mouse model, the dysplastic initial segment was found to have increased apoptosis rate, suggesting that the induced proliferative pressure by SV40 Tag expression is counter-balanced by increased rate of apoptosis. The increased apoptosis rate in the epididymides of SV40 Tag-expressing mice might be one of the mechanisms resulting to the fact that no tumor formation was detected in the epididymis. Also, in the caput epididymidis apoptosis rate was increased in GPX5Tag2 males. The rest of the epididymal regions, i.e. corpus and cauda, had normal apoptotic rate compared with WT males, suggesting a correlation between the observed alterations in the epididymal histology at the initial segment and increased apoptosis rate. However, there was no difference in the number of apoptotic cells between WT and GPX5-Tag2 testes in any of the spermatogenic stages studied, and hence, the reason for the reduced testis weight with qualitatively normal spermatogenesis remains to be explored. No change in serum testosterone and LH levels and only a mild change in the FSH and inhibin B at the old adult age was detected in the GPX5-Tag2 males, further suggesting that the infertility found in the mice was not due to hormonal factors but rather to a nonhormonal effect in the epididymis. The most obvious feature of the epididymal sperm in the GPX5-Tag2 mice was the angulated flagellum that was maintained while the sperm were swimming and did not hinder motility significantly. From the fixed epididymal preparations, it was clear that over 80% of the spermatozoa produced by the testes were straight in shape and the bending of the flagellum occurred gradually within the epididymal canal and especially as sperm entered the cauda epididymidis. By comparison, the c-ros knockout mice, whose in vivo infertility is due to failure of sperm in the uterus to negotiate the oviduct because of the hairpin forms of the sperm, exhibit much less extensive tail angulation within the epididymis, although the angulation is exaggerated upon suspension in culture medium (32, 36). Sperm angulation is a response to osmotic swelling, which forces the tail to bend at the enlarged cytoplasmic droplet, affecting an increase in cell volume without unsustainable stretching of the cell membrane. It has been shown that sperm maturation in the epididymis includes changes in properties of cell volume regulation (32), but it is not known whether the process involves modification of sperm membrane transport,



2613

Table 3. Genes Analysed from the Epididymis of GPX5-Tag2 Mice: Primer Pairs Used in RT-PCR and Radioactive Probe Generation, and Specific RT-PCR Conditions Gene

GPX5 Cysteine-rich secretory protein-1 (CRISP-1) Mouse epididymal protein 9 (MEP9) mEP17 Mouse epididymal 1 (ME1) Human epididymal 5 (HE5) HE6 CRES Murine epididymal retinoic acid-binding protein (mE-RABP)

5⬘-cttctagccagctatgtg-3⬘ 5⬘-gtactggattgtcagaccg-3⬘ 5⬘-gcttgtgctgttcttcttgg-3⬘ 5⬘-ccagatggagaaaccattcg-3⬘ 5⬘-ttagcagtggcactgtcctc-3⬘ 5⬘-ggttatacttcttgcggaagg-3⬘ 5⬘-tgaccaaaaactggtgctga-3⬘ 5⬘-ggctcatcctcgttctcaag-3⬘ 5⬘-ttccctattcctgagcctga-3⬘ 5⬘-tccatttcgggtctgatttc-3⬘ 5⬘-agtggcctgcagactgtcct-3⬘ 5⬘-tgggtcgatcccgtgttatt-3⬘ 5⬘-catccgaaaatacatcc-3⬘ 5⬘-tgaaggcacacgtctcc-3⬘ 5⬘-ggccagagtgaggaaggag-3⬘ 5⬘-tcatggcaaaccaaacacac-3⬘ 5⬘-gtttttaggcttctggtatga-3⬘

Annealing Temperature (C)

Cycles

55

13

M68896

56

22

NM_009638

55

20

BC008169

55

19

AF082221

56

15

AB021289

55

13

NM_013706

47

24

X81892a

60

15

AF091503

56

13

U68381

GenBank No.

c-ros

5⬘-ctgatattctggtgaccttgta-3⬘ 5⬘-tcacagatctacaaccttacacct-3⬘ 5⬘-agcaaccagaaatatcccaacta-3⬘

59

20

X81650

Ion transporters Sodium bicarbonate cotransporter 3 (NBC-3)

5⬘-aagtgaactgggcaaacctg-3⬘

58

24

AF224508

58

24

AF255774

58

26

NM_018760

58

23

AF139193

47

23

AF139194

64

17

X03672

Anion exhanger 2 (AE-2) Anion exhanger 4 (AE-4) Na⫹/H⫹ exhanger 2 (NHE-2) Na⫹/H⫹ exhanger 3 (NHE-3)

␤-Actin a

Sequence of Primers Pairs

5⬘-acacattgggaacctgaagc-3⬘ 5⬘-atgcggatgcacctgttc-3⬘ 5⬘-catccacaccttcacactcg-3⬘ 5⬘-gcctactgaagcctgacctg-3⬘ 5⬘-gaagtcaacctccccaacaa-3⬘ 5⬘-cttctgattcgggaaaacca-3⬘ 5⬘-atcaggatctccttggcttg-3⬘ 5⬘-cttcattcgctccccaagta-3⬘ 5⬘-tcattgggtgttagctcgtg-3⬘ 5⬘-cgtgggccgccctaggcacca-3⬘ 5⬘-ttggccttagggttcaggggg-3⬘

Primers modified from human sequence: catccgaaaAtacatcc/tgaaggcacacAtctcc.

sperm osmolytes, or osmolarity of the luminal milieu. The latter is much higher than the osmolarities of fluids in the rete testis or the female tract. The in situ angulation of the epididymal sperm in the GPX5-Tag2 mice might be related to the observed decrease in osmolality of the cauda epididymidal fluid, which could be a result of epididymal malfunction. Elimination of membrane restraint by a detergent allows angulated sperm from the c-ros knockout mice to straighten out (36). By contrast, such treatment reduced, but did not abolish, hairpin forms from GPX5-Tag2 sperm found in situ. This failure to straighten out may reflect prolonged fixation of the bent tail by thiol oxidation within the epididymis because incubation in a thiol-reducing agent induced a further reduction in hairpin forms in sperm both before and after demembranation. The tail angulation may well explain the lack of fertilized oocytes in the oviduct and the infertility of the

GPX5-Tag2 males in vivo as well as in vitro, because similar changes in sperm morphology have been shown to be the cause of infertility in the c-ros knockout mouse (37). In the c-ros knockout mice, the initial segment is lacking, suggesting that the initial segment in particular is involved in the formation of the phenotype of sperm angulation. Unlike the c-ros knockout mice, the GPX5-Tag2 mice contain the initial segment, but its mere presence may not necessarily denote normal function and critical secretions of the epithelial cells that could be missing from the GPX5-Tag2 transgenic males. The data suggest that the lower osmotic pressure of epididymal fluid could be the cause for the bent sperm phenotype seen in GPX5-Tag2 mice. However, the expression levels of all the tested ion transporters were unchanged, but whether there is a functional defect in some of the ion transporters causing the lower osmolality of caudal fluid in GPX5-Tag2 epididymides remains to be studied.



studies have shown the importance of the retinoic acid signaling pathway for the function of epididymal epithelium, and retinoids have been shown to regulate epididymal gene expression (41), whereas HE6 has been suggested to be an orphan membrane receptor. However, the direct role of these proteins in the sperm function remains to be explored further. In conclusion, the GPX5-Tag2 is a novel mouse line with epididymal defect, and it is useful for analyzing further the role of epididymis in posttesticular sperm maturation, especially to analyze genes involved in the regulation of the posttesticular sperm maturation.

MATERIALS AND METHODS Transgene Construct

Fig. 9. Epididymal Gene Expression in the GPX5-Tag2 Epididymides A, CRES gene was not expressed in the GPX5-Tag2 initial segment. B, HE6 gene expression level was lower in GPX5Tag2 mice initial segment, caput, corpus, and cauda epididymidis compared with WT mice. C, Another initial segment specific gene, mEP17, was also missing from the GPX5-Tag2 epididymides. IS, Initial segment; CAP, caput; COR, corpus; CAU, cauda.

It has been reported previously that SV40 Tag transformation of prostatic epithelial cells was accompanied by down-regulation of the differentiated function, demonstrated by the loss of differentiation-specific secretory proteins (15). Accordingly, we suggest that the SV40 Tag expression in the GPX5-Tag2 mice results in a defect in differentiation of the initial segment. This is supported by the lack of expression of certain initial segment-specific genes, such as CRES and mEP17. By contrast, all of the tested caput-, corpus-, and cauda-specific genes were normally expressed. The only exception was HE6, for which expression levels were lower than in controls in GPX5-Tag2 mice in all of the epididymal regions. In addition to being merely markers for the lack of differentiation of initial segment, these proteins could also be partially the cause of bend sperm phenotype seen in GPX5-Tag2 sperm. CRES is a member of the cystatin superfamily of cysteine protease inhibitors, and it is suggested to be involved in the regulation of proprotein processing (38). The target proteins for CRES and its function are still unclear, but the absence of CRES from the initial segment may lead to undesirable protein processing by cysteine proteases either in the epididymal epithelium or sperm plasma membrane. mEP17 is a new member of the lipocalin superfamily suggested to transport retinoids within the epididymis (39), and HE6 belongs to the seven transmembrane-domain receptor superfamily (40). Animal

A 5-kb 5⬘-fragment of the mouse GPX5 gene, corresponding to nucleotides ⫺5012 to ⫹24 was excised from the Bluescript II KS vector (Stratagene, La Jolla, CA) using the XhoI and SpeI restriction enzymes. The fragment was ligated to the pGEM-7Zf vector (Promega Corp., Madison, WI) in front of the 2.7-kb fragment of the SV40 early genes, containing the coding sequences for the large T- and small t-antigens and a polyadenylation signal. This resulted in transgene construct GPX5-Tag (Fig. 1A). Establishment of the GPX5-Tag Transgenic Mouse Lines The GPX5-Tag transgene was cleaved from the pGEM-7Zf vector by XhoI/SacI digestions (Fig. 1A), purified by a QuickPick Electroelution Capsule Kit (QIAGEN, Vale´ ncia, CA) and Elutip DEAE-columns (Schleicher & Schu¨ ell, Dassel, Germany), and diluted to the final concentration of 2 ng/␮ml. Transgenic founder mice were generated in the genetic background of the FVB/N strain by microinjecting the DNA into pronuclei of fertilized oocytes using standard techniques. Integration of the transgene was verified by PCR screening using tail DNA isolated using the salting-out method. The PCR consisted of 30 cycles (1 min 97 C, 1.5 min 56 C, 2 min 72 C), and the following primers were used: 5⬘-primer, 5⬘CAGCTAATGGACCTTCTAGG-3⬘; 3⬘-primer, 5⬘-GCAATCGAAGCAGTAGCAATC-3⬘. The founder mice were mated with WT FVB/N mice to create specific transgenic mouse lines (GPX5-Tag1 and GPX5-Tag2). All mice were handled in accordance with the institutional animal care policies of the University of Turku (Turku, Finland). The mice were specific pathogen-free and were fed with complete pelleted chow and tap water ad libitum in a room with controlled light (12 h light, 12 h darkness) and temperature (21 ⫾ 1 C). To confirm the transgene integration and analyze the copy numbers of the integrated transgene, Southern blot analysis was performed using previously described protocols (6). RNA Analysis Total RNA from various tissues was isolated using the singlestep method, and the GPX5-Tag transgene expression in the GPX5-Tag1 and GPX5-Tag2 mouse lines was studied using Northern blot analysis and RT-PCR. In addition, the expression of several epididymal genes in the GPX5-Tag2 epididymides was studied using semiquantitative RT-PCR and Northern blot analysis. For Northern blot analysis, 20 ␮g denatured total RNA were resolved on a 1% denaturing agarose gel and transferred onto nylon membranes (Hybond-XL).


The membranes were hybridized with the [32P]␣CTP-labeled cDNA for the large T-antigen gene, 308-bp-long RT-PCR product for CRES gene, and 300- and 431-bp-long RT-PCR products for mEP17 and HE6, respectively, using standard techniques. Hybridization signals were detected by autoradiography using x-ray film (Fuji Photo Film Co., Ltd., Tokyo, Japan) or a phosphor imager (Fuji Photo Film Co., Ltd.). In the RT-PCR analysis, 1 ␮g of DNase I (Life Technologies, Inc., Paisley, Scotland, UK)-treated total RNA was reverse-transcribed (10 min, 50 C) using avian myeloblastosis virus reverse transcriptase (Promega Corp.) and amplified using Dynazyme II-polymerase (Finnzymes, Espoo, Finland) in the same reaction tube. The primers used were identical to those used for genotyping the GPX5-Tag mice. 3⬘-End labeling of RT-PCR products using digoxigenin-11dideoxy-uridine triphosphate (Boerhringer Mannheim, Basel, Switzerland) was done as described previously (42). Total RNA was DNase I-treated before RT-PCR. Primers for ␤-actin (used as an internal control) and for each tested gene, annealing temperatures, and cycle numbers are represented in Table 3. Morphological, Histological, and Immunochemical Analyses Histological evaluation of tissues of WT, GPX5-Tag1, and GPX5-Tag2 mice was made in two age groups, 50–60 d and 4 months. The mice were anesthetized by ip injection of 300–600 ␮l of 2.5% Avertin (Sigma, St. Louis, MO), blood was collected by cardiac puncture, and tissues were dissected out for macroscopic analyses and for organ weights. For histological analyses, Bouin- and paraformaldehyde-fixed paraffin sections of tissues were stained with hematoxylin/ eosin. For immunohistochemistry, paraffin sections from epididymides, testes, seminal vesicle, prostate and adrenal gland of the WT and GPX5-Tag1 males, and epididymides and testes from the GPX5-Tag2 males were boiled in 1 M urea in a microwave oven for 15 min. Samples were then immunostained with a mouse monoclonal SV40 Tag (Ab-2) antibody (Oncogene Research Products, Boston, MA; 1:100 to 1:300 dilution in PBS). The antigen-antibody complexes were visualized by biotinylated anti-mouse antibody (Vector Laboratories, Inc., Burlingame, CA) combined with avidin-horseradish peroxidase complex (ABC kit, Vector Laboratories, Inc.) and 3,3⬘-diaminobenzidine tetrahydrochloride (DAB-Plus Substrate Kit, Zymed Laboratories, Inc., San Francisco, CA). TUNEL Staining To study whether the increased apoptosis rate was the cause of decreased epididymis and testis size, TUNEL staining was performed on epididymal and testicular sections of WT and GPX5-Tag2 males. The DNA strand breaks present in paraffinembedded sections were labeled with the In-situ Cell Death Detection Kit (Roche Molecular Biochemicals, Que´ bec, Canada) according to the manufacturer’s instructions. In different parts of epididymides (initial segment, caput, corpus, cauda), the number of TUNEL-positive cells present in tubular crosssections on seven individual microscopic frames (20⫻ objective) were then calculated. The cross-sections of testicular seminiferous tubules were divided into three different groups according to the spermatogenic stages: I–VI, VII–VIII, and IX–XII; the number of apoptotic cells was calculated from each group using 5 testicular cross-sections from 3 mice, or 15 sections altogether. Hormone Measurements After blood collection, the separated serum samples were stored at ⫺70 C until assayed for hormones. Serum FSH was measured by an immunofluorometric assay for rat FSH, es-


2615

sentially as described before (43). LH of the sera was measured by a sensitive immunofluorometric assay for rat LH (Delfia; Wallac, Inc. OY, Turku, Finland) as described earlier (44). Testosterone was measured from diethyl ether extracts of the sera using RIA as described earlier (45). Serum inhibin B concentrations were measured by using inhibin-B assay kit (Oxford Bio-Innovation Ltd., Oxford, UK). Analysis of the Fertility of the GPX5-Tag Male Mice Continuous matings were performed to analyze the fertility of males at the age of 7 wk to 3 months from both GPX5-Tag1 and GPX5-Tag2 lines, and none of the mice was found to be fertile. For statistical analysis of the infertility, six males from both lines were mated with three WT females each. Copulatory plugs were sought daily, and the plugged females were followed for 3–4 wk to determine the number of litters and offspring produced by each male. For GPX5-Tag2 males, additional studies were performed to determine whether the existing spermatozoa were able to fertilize oocytes in vivo. Four GPX5-Tag2 males, at the age of 2 months, and six age-matched WT males were mated with six adult females each, and females were checked daily for copulatory plugs. The plugged females were killed, and oocytes were collected from the oviductal ampulla. The number of fertilized oocytes and the total number of oocytes in the oviducts were counted. The oocytes were furthermore incubated in KSOM Embryo Culture Medium (Specialty Media Inc., Lavallette, NJ) at 37 C overnight to analyze whether the fertilized oocytes divided normally. In vitro fertilization was performed to GPX5-Tag2 males as described previously (46). Analysis of Sperm Motility To analyze whether sperm motility was affected by metabolic substrates, we followed a previously described method (47, 48). Sperm from GPX5-Tag2 mice were released from a few loops of tubule of the cauda epididymidis into two different media; medium H contained glucose, pyruvate, and lactate as substrates, and medium G contained glucose only. The sperm were allowed to disperse for 1–2 min at 37 C in 5% CO2, and sperm suspensions were then further diluted to a concentration appropriate for motility assessment at 37 C, immediately and again after a 2.5-h incubation at 37 C. The sperm suspension was placed on a siliconized slide under an 18 ⫻ 18-mm coverslip to provide a depth of 40 ␮m. Motility was recorded by videotaping using pseudo-darkfield optics created by using a 4⫻ objective with a 40⫻ condenser ring and a 3.3⫻ photo ocular. The percentages of the motile and immotile sperm with straight or bent flagella were measured using phase contrast optics and a 20⫻ objective. The kinematic parameters of 100–200 motile sperm per sample were measured by analyzing the videotapes at 25 Hz for 30 frames by the Hamilton-Thorne HTM-C 10.6 system (Beverly, MA). Analyzing Sperm Shape The entire epididymis of GPX5-Tag2 mice was immerse-fixed overnight in 5% glutaraldehyde. The organs were divided into seven regions: proximal and distal caput; proximal, middle, and distal corpus; and proximal and distal cauda epididymidis. Tubule fragments were rinsed and minced in PBS before brief sonication to release individual sperm from the fixed, agglutinated clumps. Five-microliter aliquots were examined for categorization of the sperm tail as straight, slightly angulated (obtuse angle), greatly angulated (acute angle), or hairpin bends. In an additional experiment, spermatozoa were released by cutting a few tubule segments from the caput, corpus, and cauda epididymidis of WT and GPX5-Tag2 mice and transferring the luminal contents into BWW medium (Specialty



Media Inc.) containing glucose, pyruvate, and lactate and sufficient NaCl to be of the same osmolality as that of cauda epididymidal fluid from GPX5-Tag2 mice. After dispersion, 20 ␮l were either fixed in 2.5% glutaraldehyde in PBS or treated with 1% Triton X-100 in PBS for 1 min before fixation to see whether the bending was caused by membrane restraint. Other aliquots of caudal sperm were released into medium containing 1 mM DTT and incubated for 45 min at 37 C, after which time they were fixed as above. For each sample, 100 sperm were categorized for their morphology. The total percentage of angulated forms was calculated from all angulated flagella, including hairpin forms. The percentage of sperm bearing cytoplasmic droplets was calculated by adding the straight sperm bearing visible droplets to those that were angulated at the site of the droplet.

medicine, University of Turku, Kiinamyllynkatu 10, FIN-20520 Turku, Finland. E-mail: [email protected]. This work was supported by grants from the Rockefeller and Ernst-Schering Research Foundations, the Academy of Finland (project nos. 42145 and 43745), the Turku University Foundation, and the Turku Graduate School of Biomedical Sciences.

Measurement of Osmotic Pressure The osmotic pressure of 3 ␮l of undiluted cauda epididymidal contents was measured by a Vapro vapor pressure osmometer as described by Yeung et al. (36), with the exception that fluid was flushed out with a medium possessing osmolality of 430 mmol NaCl/kg water. Six samples were obtained from different GPX5-Tag2 males and 8 samples from WT males. For each sample, the mean of two measurements (taken after 10 and 15 min equilibration) was taken. In addition to calibration with standards of 100, 290, and 1000 mmol/kg as recommend by the supplier, standards of 400 mmol NaCl/kg water were measured under identical conditions before, during, and after measurements of the epididymal fluids to confirm the absence of instrument drift. Statistical Analyses Statistical analyses were performed by SigmaStat-program (version 2.0 for Windows 95, SPSS, Inc., Chicago, IL). For TUNEL results, t test or Mann-Whitney rank sum test was performed for analyzing the statistical significance (P ⬍ 0.05). For hormone results, Kruskal-Wallis one-way analysis or oneway ANOVA was performed, and in the case of statistically significant results, Dunn’s or Tukey’s test was performed for pair-wise multiple comparisons. For tissue weights, the same analyses were performed. Two-way ANOVA was used to examine the differences in the percentages of morphological forms of the sperm for each genotype, in each epididymal region separately, in the absence and presence of Triton X-100. One-way ANOVA was used to examine the effects of DTT on tail morphology of caudal GPX5-Tag2 sperm. The t test was used to compare the osmotic pressures of epididymal fluid between genotypes.

Acknowledgments We thank Prof. J. Green (University of California, San Francisco, CA) for providing the T-antigen cDNA; Dr. L. A¨ hrlundRichter (Clinical Research Center, Karolinska Institutet, Huddinge University Hospital, Sweden) for performing the in vitro fertilization; Docent J. Toppari, M.D., Ph.D., and Dr. W. Yan (Department of Physiology, University of Turku) for analyzing the testis specimens; Dr. K. Alanen (Department of Anatomy, University of Turku) for helping with histological analyses; T. Prins for help in optimizing RT-PCR conditions; N. Messer for the DNA microinjection and animal handling; M. Forsblom, B. Helle, and M. Mertanen for animal husbandry; J. Vesa for the technical assistance with the histology specimens; and T. Laiho for technical assistance with hormone measurements.

Received March 11, 2002. Accepted July 23, 2002. Address all correspondence and requests for reprints to: Matti Poutanen, Department of Physiology, Institute of Bio-

REFERENCES 1. Robaire B, Hermo L 1988 Efferent ducts, epididymis, and vas deferens: structure, functions, and their regulation. In: Knobil E, Neill J, eds. The physiology of reproduction. New York: Raven Press Ltd.; 999–1080 2. Cooper TG 1998 Epididymis. In: Neill JD, Knobil E, eds. Encyclopedia of reproduction 2. San Diego: Academic Press; 1–17 3. Orgebin-Crist M-C, Danzo BJ, Davies J 1975 Endocrine control of the development and maintenance of sperm fertilizing ability in the epididymis. In: Hamilton DW, Greep RO, eds. Handbook of physiology. Washington DC: American Physiological Society; 319–338 4. Kirchhoff C 1997 Molecular aspects of epididymal function and sperm maturation. In: Waites GMH, Frick J, Baker GWH, eds. Current advances in andrology. Bologna, Italy: Monduzzi; 253–259 5. Lareyre J-J, Thomas TZ, Zheng W-L, Kasper S, Ong DE, Orgebin-Crist M-C, Matusik RJ 1999 A 5-kilobase pair promoter fragment of the murine epididymal retinoic acid-binding protein gene drives the tissue-specific, cellspecific, and androgen-regulated expression of a foreign gene in the epididymis of transgenic mice. J Biol Chem 274:8282–8290 6. Lahti PP, Shariatmadari R, Penttinen JK, Drevet JR, Haendler B, Vierula M, Parvinen M, Huhtaniemi IT, Poutanen M 2001 Evaluation of the 5⬘-flanking regions of murine glutathione peroxidase five and cysteine-rich secretory protein-1 genes for directing transgene expression in mouse epididymis. Biol Reprod 64:1115–1121 7. Ghyselinck NB, Dufaure I, Lareyre JJ, Rigaudiere N, Mattei MG, Dufaure JP 1993 Structural organization and regulation of the gene for the androgen-dependent glutathione peroxidase-like protein specific to the mouse epididymis. Mol Endocrinol 7:258–272 8. Ghyselinck NB, Jimenez C, Dufaure JP 1991 Sequence homology of androgen-regulated epididymal proteins with glutathione peroxidase in mice. J Reprod Fertil 93:461–466 9. Ghyselinck NB, Jimenez C, Courty Y, Dufaure JP 1989 Androgen-dependent messenger RNA(s) related to secretory proteins in the mouse epididymis. J Reprod Fertil 85:631–639 10. Jimenez C, Ghyselinck NB, Depeiges A, Dufaure J-P 1990 Immunochemical localization and association with spermatozoa of androgen-regulated proteins of MR 24000 secreted by the mouse epididymis. Biol Cell 68:171–174 11. Vernet P, Faure J, Dufaure J-P, Drevet JR 1997 Tissue and developmental distribution, dependence upon testicular factors and attachment to spermatozoa of GPX5, a murine epididymis-specific glutathione peroxidase. Mol Reprod Dev 47:87–98 12. Vernet P, Rigaudie´ re N, Ghyselinck N, Dufaure JP, Drevet JR 1996 In vitro expression of a mouse tissue specific glutathione-peroxidase-like protein lacking the selenocysteine can protect stably transfected mammalian cells against oxidative damage. Biochem Cell Biol 74:125–131 13. Vernet P, Rock E, Mazur A, Rayssiguier Y, Dufaure J-P, Drevet JR 1999 Selenium-independent epididymisrestricted glutathione peroxidase 5 protein (GPX5) can back up failing Se-dependent GPXs in mice subjected to selenium deficiency. Mol Reprod Dev 54:362–370


14. Kriegler M, Perez CF, Hardy C, Botchan M 1984 Transformation mediated by the SV40 T antigens: separation of the overlapping SV40 early genes with a retroviral vector. Cell 38:483–491 15. Kasper S, Sheppard PC, Yan Y, Pettigrew N, Borowsky AD, Prins GS, Dodd JG, Duckworth ML, Matusik RJ 1998 Development, progression, and androgen-dependence of prostate tumors in probasin-large T antigen transgenic mice: a model for prostate cancer. Lab Invest 78: 319–333 16. Kananen K, Markkula M, El-Hefnawy T, Zhang F-P, Paukku T, Su J-GJ, Hsueh AJW, Huhtaniemi I 1996 The mouse inhibin ␣-subunit promoter directs SV40 T-antigen to Leydig cells in transgenic mice. Mol Cell Endocrinol 119:135–146 17. Cairns LA, Crotta S, Minuzzo M, Ricciardi-Castagnoli P, Pozzi L, Ottolenghi S 1997 Immortalization of neuroendocrine cells from adrenal tumors arising in SV40 Ttransgenic mice. Oncogene 14:3093–3098 18. Livingston DM, Bradley MK 1987 The simian virus 40 large T antigen. A lot packed into a little. Mol Biol Med 4:63–80 19. Ludlow JW 1993 Interactions between SV40 large-tumor antigen and the growth suppressor proteins pRB and p53. FASEB J 7:866–871 20. Tzeng YJ, Zimmermann C, Guhl E, Berg B, Avantaggiati ML, Graessmann A 1998 SV40 T/t-antigen induces premature mammary gland involution by apoptosis and selects for p53 missense mutation in mammary tumors. Oncogene 16:2103–2114 21. Faure J, Ghyselinck N-B, Jimenez C, Dufaure JP 1991 Specific distribution of messenger ribonucleic acids for 24-kilodalton proteins in the mouse epididymis as revealed by in situ hybridization: developmental expression and regulation in the adult. Biol Reprod 44:13–22 22. Dufaure J-P, Lareyre J-J, Schwaab V, Mattei M-G, Drevet JR 1996 Structural organization, chromosomal localization, expression and phylogenetic evaluation of mouse glutathione peroxidase encoding genes. Acad Sci Paris 319:559–568 23. Muller WJ, Sinn E, Pattengale PK, Wallace R, Leder P 1988 Single-step induction of mammary adenocarcinoma in transgenic mice bearing the activated c-neu oncogene. Cell 54:105–115 24. Bouchard L, Lamarre L, Tremblay PJ, Jolicoeur P 1989 Stochastic appearance of mammary tumors in transgenic mice carrying the MMTV/c-neu oncogene. Cell 57:931–936 25. Gilbert E, Morel A, Tulliez M, Maunoury R, Terzi F, Miquerol L, Kahn A 1997 In vivo effects of activated h-ras oncogene expressed in the liver and in urogenital tissues. Int J Cancer 73:749–756 26. Behringer RR, Peschon JJ, Messing A, Gartside CL, Hauschka SD, Palmiter RD, Brinster RL 1988 Heart and bone tumors in transgenic mice. Proc Natl Acad Sci USA 85:2648–2652 27. Nayernia K, Samani AA, Klaproth S, Engel W 1998 Haploid male germ cells show no susceptibility to transformation by simian virus 40 large tumour antigen in transgenic mice. Cell Biol Int 22:437–443 28. Suzuki F 1982 Microvasculature of the mouse testis and excurrent duct system. Am J Anat 163:309–325 29. LeBarr DK, Blecher SR 1987 Epididymides of sex-reversed XX mice lack the initial segment. Dev Genet 7:109–116 30. Elliott PR 1965 The metabolism of homogenates of the mouse epididymis. J Cell Comp Physiol 66:281–292 31. Sonnenberg-Rietmacher E, Walter B, Riethmacher D, Go¨ decke S, Birchmeier C 1996 The c-ros tyrosine kinase receptor controls regionalization and differentiation of epithelial cells in the epididymis. Genes Dev 10: 1184–1193 32. Yeung CH, Sonnenberg-Riethmacher E, Cooper TG 1998 Receptor tyrosine kinase c-ros knockout mice as a


33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

2617

model for the study of epididymal regulation of sperm function. J Reprod Fertil Suppl 53:137–147 Allemand I, Grimber G, Kornprobst M, Bennoun M, Molina T, Briand P, Joulin V 1995 Compensatory apoptosis in response to SV40 large T antigen expression in the liver. Oncogene 11:2583–2590 Shibata M-A, Maroulakou IG, Jorcyk CL, Gold LG, Ward JM, Green JE 1996 p53-independent apoptosis during mammary tumor progression in C3(1)/SV40 large T antigen transgenic mice: suppression of apoptosis during the transition from preneoplasia to carcinoma. Cancer Res 56:2998–3003 Naik P, Karrim J, Hanahan D 1996 The rise and fall of apoptosis during multistage tumorigenesis: downmodulation contributes to tumor progression from angiogenic progenitors. Gen Dev 10:2105–2116 Yeung C-H, Sonnenberg-Riethmacher E, Cooper TG 1999 Infertile spermatozoa of c-ros tyrosine kinase receptor knockout mice show flagellar angulation and maturational defects in cell volume regulatory mechanisms. Biol Reprod 61:1062–1069 Yeung C-H, Wagenfeld A, Nieschlag E, Cooper TG 2000 The cause of infertility of male c-ros tyrosine kinase receptor knockout mice. Biol Reprod 63:612–618 Cornwall GA, Hsia N, Sutton HG 1999 Structure, alternative splicing and chromosomal localization of the cystatin-related epididymal spermatogenic gene. Biochem J 340:85–93 Lareyre J-J, Winfrey VP, Kasper S, Ong DE, Matusik RJ, Olson GE, Orgebin-Crist M-C 2001 Gene duplication gives rise to a new 17-kilodalton lipocalin that shows epididymal region-specific expression and testicular factor(s) regulation. Endocrinology 142:1296–1308 Osterhoff C, Ivell R, Kirchhoff C 1997 Cloning of a human epididymis-specific mRNA, HE6, encoding a novel member of the seven transmembrane-domain receptor superfamily. DNA Cell Biol 16:379–389 Cornwall GA, Lareyre J-J, Matusik RJ, Hinton BT, Orgebin-Crist M-C 2002 Gene expression and epididymal function. In: Robaire B, Hinton BT, eds. The epididymis from molecules to clinical practice. New York: Kluwer Academic/Plenum Publishers; 169–199 Zheng H, Wei Y, Toppari J, Ha¨ rko¨ nen P 2000 Improved nonradioactive RT-PCR method for relative quantification of mRNA. Biotechniques 28:832–834 Hakola K, Van der Boogaart P, Mulders J, de Leeuw R, Schoonen W, Van Heyst J, Swolfs A, Van Casteren J, Huhtaniemi I, Kloosterboer H 1997 Recombinant rat follicle-stimulating hormone: production by Chinese hamster ovary cells, purification and functional characterization. Mol Cell Endocrinol 127:59–69 Haavisto A-M, Pettersson K, Bergendahl M, Perheentupa A, Roser J, Huhtaniemi I 1993 A supersensitive immunofluorometric assay for rat luteinizing hormone. Endocrinology 132:1687–1691 Huhtaniemi I, Nikula H, Rannikko S 1985 Treatment of prostatic cancer patients with a gonadotropin-releasing hormone agonist analog: acute and long term effects on endocrine functions of testis tissue. J Clin Endocrinol Metab 61:698–704 Thornton CE, Brown SDM, Glenister PH 1999 Large numbers of mice established by in vitro fertilization with cryopreserved spermatozoa: implications and applications for genetic resource banks, mutagenesis screens, and mouse backcrosses. Mamm Genome 10:987–992 Yeung CH, Oberla¨ nder G, Cooper TG 1992 Characterization of the motility of maturing rat spermatozoa by computer-aided objective measurement. J Reprod Fertil 96:427–441 Cooper TG, Yeung CH, Skupin R, Haufe G 1997 Antifertility potential of ornidazole analogues in rats. J Androl 18:431–438